[KLUG Advocacy] A [little] more on the SCO suit

[Robert G. Brown]

I had quoted:
>>>"It is not possible for Linux to rapidly reach UNIX performance standards
>>>for complete enterprise functionality without the misappropriation of UNIX
>>>code, methods or concepts to achieve such performance, and
>>>coordination by a larger developer, such as IBM."

Adam Williams REsponded:
>This statement borders on "What!?".  The *ONLY* way to achieve
>"enterprise functionality" is to appropriate chunks of the sacred and
>holy one-only-true UNIX code from ~20 years ago?  Come on!  
I quoted it becasue I thought it bordered on the comical, actually. I
wasn't agreeing with it.

>Do you think Microsoft or Novel would agree with that statement?
They might, officially. Mostly, it tends to fit the "Cathedral" model of
software development. Since we know the power of the "bazaar", we know 
that this is just silly.

>And is "enterprise functionality" the same now as 5, 10, 15 years ago?  
No, nor 6 months ago.

>This seems like a reference to some teleological constant; I can't 
>imagine an argument with such a premise would survive serious scrutiny
>for long, even by someone with a non-technical background.
Seems about right. We you expecting me to DEFEND that quote? You're going
the need a LOT of patience! :)

John Bridleman added:
>>Yea, whatever. Seems like they would have to give specific code examples. 
>>The irony here is that Caldera, um I mean SCO, has a Linux distro.
You would think that Cal..SCO would benefit from that, and be happy. 
It hasn't turned out that way.

Adam answered...
>Right, but they are challenging IBM's right to extend Linux
>functionality with UNIX code, mostly in reference to glibc from the gist
>I've taken from the article and my knowledge of the areas in which IBM
>fund developers; threading models, etc...  Feature enhancements like
>JFS, etc.. don't really relate back to UNIX at all but are actually
>trademarked and patented by IBM itself.  Of course, glibc is not
>confined to Linux at all, which makes the argument even more dicey.

Indeed, enter the noted IBM assiduous attiude about intellectual property.
IMO they are VERY unlikely to have messed this up. If it's their stuff,
then it's their stuff to free, and Cal..SCO can like it, or lump it.

>Maybe one of the reasons Linux is replacing UNIX is because it isn't
>UNIX; that is, it has the flexibility to not be UNIX-like when being
>UNIX-like (static /dev for example) no longer makes good sense.  Linux
>is the next evolutionary step (vs. the divergent Windows and Novell). 
Perhaps. I think there's a stronger reason...

>SCO simply hasn't progressed, and IBM and others would rather not pay
>$$,$$$,$$$,$$$ to progress their products.  I smell a thick frothy stew
>of sour grapes and last-ditch-effort.
Yes, that's all true, but there is another compelling reason for IBM and
other hardware makers (e.g., HPaq) to invest in, and market on Linux.

Let's face it, hardware makers do NOT really want to be in the OS biz.
it is, by itself, a huge investment, and again (by itself) the least 
profitable segmant of their activity. They'de much rather outsource
it. This is one basis for why Microsoft has been such as success (there
are other factors, topic for another thread, or maybe we had those
threads, see the archives), but more importantly, we might want to look
at the kind of company IBM is becoming....

They are going to be making hardware (desgining it, developing it, packag-
ing it, etc.) for a long, long time. The other segment they want to go into
is the high-margin end, consulting and value-added software development.
That, folks, is where the money is. The OS costs a pile to create and 
maintain, and it's really, in marketing terms, part of the machine, and so
tends to add to basic sales costs without really adding a lot of (perceived)
value.

So the best thing to do is to get the OS (to a great degree) out of the 
picture. The problem is that the last couple of attempts were failures,
or harmful, or maybe just too expensive, or perhaps more or less outmoded.
Example: One time, they let that pesky kid from Washington state in the 
         door, and look what he did... almost ruined OS/2! :)
          
Look what Linux offers to IBM (or HPaq, or others, who might not be quite
bright enough to go for it just yet)... you get a complete OS, lots of re-
cent applications and tools, a lot of fairly open, or poor documentation
(but everyone's used to that, hire a few hundred people to rewrite, improve, 
or interpret the documentation, they're way ahead), and TONS of people who
are busy making it all better who are NOT ON THE PAYROLL, which sure effects
the outlay for the OS.

The OS comes cheap, is of very high quality, and has few admin costs.
IBM can then concentrate on high-value-added applications, and they can 
salt the OS with features they need... in effect, they have all the advan-
tages of their own OS, but rather few of the disadvantages. 

I wonder how much of the touted billion IBM spent on Linux went into base
level distro stuff (glibc, kernel, etc.), system-level apps (SAMBA, Apache,
LDAP, etc.), and how much went into value-added (proprietary) apps...

>>it amazed me just how much Darl McBride, the President and CEO, didn't
>>know when asked relatively simple questions like how much revenue is
>>IBM to SCO. Seems like he'd know that before he sued them. He also kept
>>mentioning that these questions will all be answered as the lawsuit
>>unfolds.
Looks like shooting from the hip, counting on a very extended discovery
process, or something similar. Maybe a combination.

>>I seriously doubt this will ever see a court room. I look for
>>IBM to either buy UNIX or dump AIX. 
Neither is a nesseary outcome. At some point, IBM will write a check, for
rather (or way) less money than SCO wants, and it will be over. SCO can then
use the money to settle outstanding debts and otherwise spin down the
operation.

>IBM is about at likely to dump AIX in the near-term as Bill is to dump
>Windows.  They have WAY to many HUGE customers depending on it.
They will wean people away from AIX, over time. Linux will eventually
replace it. IBM knows better than to do it fast, or even completely.
What they're looking for is to reduce their AIX-related expenses to a 
minimum, and probably migrate more and more of the "easy" clients, and
then slowly move on the others.
							Regards,
							---> RGB <---